CN-121998280-A - Day-ahead scheduling plan making method based on quotient gradient system assistance and considering frequency security constraint

CN121998280ACN 121998280 ACN121998280 ACN 121998280ACN-121998280-A

Abstract

The invention relates to a business gradient system assisted day-ahead scheduling plan making method considering frequency safety, which is implemented by collecting and processing power grid data; and calculating the constructed economic dispatch optimization model based on a quotient gradient system auxiliary interior point method to obtain a daily dispatch plan. The method ensures that the scheduling result generated by model optimization solving can give consideration to economic efficiency and safety margin, and effectively improves the applicability of actual power grid production planning. In addition, by adopting a mode of combining the integral of a commercial gradient system and the accurate solution of an interior point method, the algorithm is ensured to be reliably converged on the premise of the feasibility of the problem, and an accurate diagnosis result is provided for operators to refer to when the problem is not feasible.

Inventors

MA JIAN
LI TENGMU
HUANG HE
LI JIALU
HUANG MEIYING
CHEN YIPING
LEI AOYU
XIAO LIANG
HUANG FANGNENG
JU YICHENG

Assignees

中国南方电网有限责任公司

Dates

Publication Date: 20260508
Application Date: 20251203

Claims (5)

1. A day-ahead scheduling plan making method based on quotient gradient system assistance and considering frequency security constraint is characterized by comprising the following steps: step 1, collecting and processing power grid data; Step 2, constructing an objective function and a constraint set of a daily economic dispatch optimization model according to the processed power grid data; And 3, calculating the day-ahead economic dispatch optimization model constructed in the step 2 based on a quotient gradient system auxiliary interior point method strategy to generate a day-ahead dispatch plan, wherein the day-ahead economic dispatch plan comprises an active and reactive dispatch plan, an energy storage charge-discharge plan and a new energy output plan of various units in a period.
2. The method for planning a day-ahead dispatch plan based on a commercial gradient system as set forth in claim 1, wherein the grid data in step 1 includes network topology and line parameters, period load and new energy output prediction data, and system operating conditions, and deletion, mean value replacement and filling are adopted to process missing values and outliers in the data.
3. The method for planning a day-ahead schedule based on quotient gradient system assistance and accounting for frequency security constraints of claim 1, wherein the objective function in step 2 is to minimize conventional unit operation costs and new energy power-off penalty costs, and the operation costs include adjustment costs including conventional unit fuel costs and automatic power generation control AGC units: ; Wherein, the As a set of time indices, And Respectively a conventional unit and an AGC unit index set, Is a set of new energy units, For the active output of the conventional unit g at time t, And The power setting and power allocation factor of AGC unit j at time t are respectively, And The power forecast value and the actual scheduling value of the new energy unit k at the time t are used, As a function of the cost of a conventional unit, For the adjustment cost function of the AGC unit, The electricity discarding cost is the unit new energy.
4. The day-ahead scheduling planning method based on the quotient gradient system assistance and considering the frequency security constraint, which is characterized in that the constraints of the optimization model in the step 2 comprise power balance constraints, frequency constraints, voltage security constraints, line thermal constraints, unit output constraints, climbing constraints, rotation standby constraints, energy storage operation constraints and new energy output constraints.
5. The day-ahead scheduling planning method based on the quotient gradient system assistance and considering the frequency security constraint of claim 1, wherein the specific implementation method of the step 3 is that a constraint set is constructed according to the constraint in the step 2: ; Wherein each sub-constraint set comprises a power balancing constraint set: ; Wherein, the For the active output of the conventional unit i at time t under one adjustment, To once adjust the voltage amplitude of the lower node i at time t, To once adjust the voltage amplitude of the lower node j at time t, And For the conductance and susceptance of branch i-j, For the phase angle difference of nodes i and j at time t, For reactive output of the conventional unit i at time t under one adjustment, For the active output of the conventional unit i at the time t under the secondary regulation, For the second adjustment of the voltage amplitude of the lower node i at time t, To secondarily adjust the voltage amplitude of the lower node j at time t, Reactive power output of the conventional unit i at a time t is regulated for the second time; Frequency safety constraint set: ; Wherein, the In order to adjust the full-network frequency at one time, For the nominal frequency of the system, 、 The upper limit and the lower limit of the system frequency are sequentially set; voltage safety constraint set: ; Wherein, the And The lower limit and the upper limit of the voltage amplitude of the node i are respectively; Line thermal constraint set: ; Wherein, the For the set of system lines, And The active power and reactive power delivered from node i to node j via line i under one regulation respectively, And The active power and the reactive power are reversely transmitted to the node i from the node j through the line l under one regulation, Power transfer capacity for line l; Set of unit output constraints: ; Wherein, the And Respectively adjusting the active and reactive output dispatching values of the lower unit g at the time t for one time, And Respectively adjusting the active and reactive output dispatching values of the generator set g at the time t for the second time, And the unit adjusting power under one-time adjustment of the conventional unit g, For the full network frequency after one adjustment, For the nominal frequency of the system, And The upper limit and the lower limit of the active output of the conventional unit g are respectively, And The upper limit and the lower limit of the reactive output of the conventional unit g are respectively, The power of the AGC unit g is allocated a factor, Unbalanced power of the whole network at the time t caused by prediction errors and real-time fluctuation of the supply and demand sides; set of constraint on climbing: ; Wherein, the And Is the lower limit and the upper limit of the climbing power of the unit g, Is the set value of the active output of the unit g, Active output of the unit g in the t-1 period; Rotating the spare constraint set: ; Wherein, the And For up/down rotation provided by the group g at time t, And Representing the minimum upward/downward rotation needed by the power grid for standby respectively; energy storage operation constraint set: ; Wherein, the For the electric energy value of the energy storage unit e at the time t, For the electric energy value of the energy storage unit e at the time t, And The charge/discharge efficiency of the energy storage unit e, For the time interval of each period of time, And Respectively the minimum/maximum charge of the energy storage unit e, And The maximum/small charge and discharge power of the energy storage unit e are respectively, And Respectively charging and discharging power of the energy storage unit e at t time, And The electric energy values of the energy storage unit e at the initial time and the final time are respectively; New energy output constraint set: ; Wherein, the , The output dispatching value and the predicted value of the new energy unit r at the time t are obtained; Wherein, the original problem decision variables Using constraint set vectors And jacobian matrix thereof A quotient gradient system is constructed: ; Wherein, the For deciding the change rate of the variable x to time, integrating the power system by adopting the ode15s, finally converging to a conventional stable balance point, wherein a one-to-one correspondence exists between the balance point and a feasible solution, taking the feasible solution as an initial solution, and accurately solving the problem by adopting IPOPT.

Description

Day-ahead scheduling plan making method based on quotient gradient system assistance and considering frequency security constraint Technical Field The invention belongs to the technical field of power system dispatching, and particularly relates to a day-ahead dispatching plan making method based on quotient gradient system assistance and considering frequency safety constraint. Background When the electric power industry makes the combination of the unit before the day and the economic dispatch plan, the linear tide equation is widely adopted as the approximation of the nonlinear tide equation so as to ensure the reliable convergence of the algorithm. Among them, the direct current power flow model (Direct Current Optimal Power Flow, DCOPF) method is most widely used. The method adopts a linear main flow equation to replace the original nonlinear flow equation so as to realize the linear approximation of the optimal flow model. However, the direct current flow equation assumes that the active and phase angle are linear, resulting in some loss of accuracy. According to measurement and calculation, the direct current trend approximately causes the loss of billions of yuan for the electric industry every year. In addition, because the reactive power and voltage information is not considered, the obtained unit dispatching result cannot embody an active transmission bottleneck caused by reactive power and voltage constraint limitation in a power grid, and the safety of the unit dispatching result applied to actual production operation cannot be ensured. Therefore, an effective optimal power flow modeling and solving method is needed in the power grid. In a day-ahead scheduling model based on alternating optimal power flow, nonlinearities of the power flow equation result in the solved optimal power flow problem being a non-convex optimization problem. Non-convex optimization solutions have the following bottlenecks: 1) The algorithm cannot guarantee reliable convergence, and 2) even if the algorithm converges, the global optimality of the solution cannot be guaranteed. In fact, the difficulty in solving the optimal power flow problem is extremely high, and when the optimal power flow problem is solved by using a numerical method such as an interior point method, if the algorithm diverges, it is difficult to determine whether an optimal solution cannot be found for numerical reasons or the original problem itself does not have an optimal solution. In the aspect of day-ahead scheduling considering safety constraint, a great deal of research, operation and optimization problems are developed and have certain achievements at present, but the safety constraint and network transmission capacity are not considered enough, frequency constraint is considered in a fresh way, particularly, the current renewable energy source is accessed in a large scale, the influence of source-load bilateral volatility on the system frequency safety is increasingly obvious, and therefore, the frequency constraint is considered more necessary when a day-ahead scheduling plan is formulated. Disclosure of Invention The invention aims to overcome the defects of the prior art, provides a day-ahead scheduling plan making method based on quotient gradient system assistance and considering frequency safety constraint, and covers power balance constraint, frequency safety constraint, voltage safety constraint, transmission limit constraint, unit output constraint, unit climbing constraint, rotation reserve constraint, energy storage operation constraint and new energy output constraint, so that a scheduling plan with economic benefit and safety margin can be reliably solved under the condition that the problem is feasible. The invention solves the technical problems by adopting the following technical scheme: a day-ahead scheduling plan making method based on quotient gradient system assistance and considering frequency security constraint comprises the following steps: and step 1, collecting and processing power grid data. Network topology and line parameters, 24/96 period load, new energy output prediction data and system operation conditions need to be obtained from a power grid data platform, and deletion values and abnormal values in the data are processed in modes of deleting, mean value replacement, filling and the like. And 2, constructing an objective function and a constraint set of a daily economic dispatch optimization model according to the processed power grid data. The operation cost comprises the fuel cost of a conventional unit and the adjustment cost of an Automatic Generation Control (AGC) unit (1) Wherein, the As a set of time indices,AndRespectively a conventional unit (comprising an AGC unit and a non-AGC unit) and an AGC unit index set,Is a set of new energy units,For the active output of the conventional unit i at time t,AndThe power setting and power allocation factor of AGC unit j at time t are respectively,AndThe power